The format of the standard rapid test lateral flow device has remained unchanged for around ten years. Typically, the device will comprise a nitrocellulose strip. Sample is applied to an application zone, from which it flows by capillary action through a zone containing a visibly-labelled antibody specific for the analyte in question. Free and bound label continue to migrate to a capture zone, where immobilised antibody specific for the analyte binds the analyte-label complex. Free label (unbound antibody) continues to migrate, leaving an analyte-specific signal in the capture zone. These types of lateral flow device are disclosed in, for example, EP-A-0284232. Numerous variations to the basic assay have been described, including those in WO92/12428, EP-A-0613005, WO97/06439, and U.S. Pat. No. 5,741,662.
In all cases, however, capture of the analyte-label complex is mediated by an immobilised reagent, which is typically an antibody that is specific for the analyte. This is unsatisfactory in many respects.
Firstly, manufacturing quality control is difficult. The solid phase capture membrane is typically made from nitrocellulose, and antibodies are applied to the membrane directly. Nitrocellulose manufacture is not, however, homogeneous. Quality control of the solid phase antibody is therefore limited to testing a statistical sample of devices from the same, but heterogeneous, batch, and assuming that the whole batch will perform within specific tolerances. It is well known, however, that membranes vary considerably, even within a single batch or lot number.
Secondly, they are relatively cumbersome to manufacture. The application of immobilised antibody to the strip requires a separate step from the application of the mobile labelled antibody. The capture antibody can be sprayed directly onto the nitrocellulose strip, but the label antibody has to be soaked into material which is subsequently attached to the nitrocellulose strip, with an overlap to ensure capillary flow.
Thirdly, antibody is immobilised by spraying a solution onto a membrane. Some of the antibody does not bind to the membrane strongly, however, and some remains loosely associated with immobilised antibody. This semi-bound or unbound antibody can become mobile when the solvent front passes over it, resulting in lower binding of label at the detection zone. If the device includes a control line, this will capture the additional label which should have been captured at the detection zone. Tests that rely on a comparison of colour intensity between control and detection lines, such as ovulation prediction kits, may therefore give false results. Furthermore, application by spraying inevitably leads to diffusion into the membrane, leading to a more diffuse and less focused detection signal.
Fourthly, the sensitivity of the devices is limited by their format. Analyte and labelled-antibody react as they migrate through the membrane, and flow rates are therefore adjusted to enable the labelled-antibody to flow at the solvent front in order to maximise the amount of time in which the analyte-label complex can form. The complex passes over the capture antibody for a short time, however, thus imposing constraints on the design of the test and its performance characteristics. The short reaction time decreases sensitivity, and also means that high affinity capture antibodies are required.
Finally, the shelf-life of these test devices is often limited by the collapse of the immobilised capture antibody onto the membrane over time.
These shortcomings in the prior art devices are addressed by the present invention, which does not use immobilised antibody to capture an analyte-label complex.